I assume the numbers you are showing below are expected frequencies for cells and that you are wishing to use a chi-square test. Right so far? If do you have 10 levels of a single categorical variable, or is it a 5x2 table?
If it's a 5x2 table, you don't actually need all of the expected frequencies to be equal to 5 or more. According to Yates, Moore & McCabe (1999, p. 734), the requirement for contingency tables larger than 2x2 is that, "No more than 20% of the expected counts are less than 5 and all individual expected counts are 1 or greater".
But if you have a 5x2 table, you don't meet even that more lenient condition--60% of your expected frequencies are < 5.
To combine categories, do something like the following. Suppose you want to combine categories 1 and 2 and 3-5 for the variable with 5 categories.
TEMPORARY.
RECODE CatVar5 (1 2 = 1) (3 4 5 = 2).
CROSSTABS CatVar5 by CatVar2 /Stat=chisqr.
This would reduce it to a 2x2 table, in which case you could actually use the N-1 chi-square. See the following link for details:
http://www.iancampbell.co.uk/twobytwo/background.htmHTH.
shihuachen wrote
hi, i want to clapse some Expcted frequency below 5 using spss syntax. ie, there are 10 counts (cases) in a count varible ,this is 1, 2, 3.5, 5 ,6 ,7, 7. 4, 2, 1. i want to use syntax conserned command aggregate to selectively to clapse counts neibouring to ensure the counts upto 5 for chi-square test. please help me for syntax problem. many thanks.
shihuachen
--
Bruce Weaver
bweaver@lakeheadu.ca
http://sites.google.com/a/lakeheadu.ca/bweaver/"When all else fails, RTFM."
PLEASE NOTE THE FOLLOWING:
1. My Hotmail account is not monitored regularly. To send me an e-mail, please use the address shown above.
2. The SPSSX Discussion forum on Nabble is no longer linked to the SPSSX-L listserv administered by UGA (
https://listserv.uga.edu/).
Locked
